Problems of Proof

• Author: Jamie Cassels - Craig Jones
• Pages: 249-272

5. Problems of Proof

A. THE USE OF STATISTICAL EVIDENCE IN MASS TORT

As will be clear from the discussion in Chapter Four, we suggest that legal-

ly and morally, there is a strong case for the use of probabilistic analysis in

mass tort claims. Legally, because statistics can indicate that the defendant’s

activity was a significant contributing factor (i.e., beyond de minimus) to the

harm in the aggregate population, even if not in any individual case. Moral-

ly, because indications of causation within a population, even without proof

of cause in any case, remove the possibility that a truly “innocent” defendant

is being held responsible despite having caused no harm at all.

Yet there is a very real concern, captured by Professor Brennan, that

courts may be unable to divorce themselves wholly from particularistic

(Brennan calls it corpuscularian) thought, particularly in individual cases:

“[I]t is not enough simply to say that courts should adopt probabilistic rea-

soning. They must be instructed. But given the importance of the moral

concept of individual responsibility in tort law, we can expect courts to

accommodate only so much probabilistic reasoning.”1

Inescapably, probabilistic proof in mass tort claims means the admis-

sion of and reliance upon statistical evidence at every step of the legal

249

1 T.A. Brennan, “Causal Chains and Statistical Links: The Role of Scientific Uncertainty

in Hazardous-Substance Litigation” (1987–88) 73 Cornell L. Rev. 469 at 491.

THE LAW OF LARGE-SCALE CLAIMS250

analysis, from causation to damages.2Various judicial justifications for this

approach have been made,3though the most frequently stated one seems

simply to be based on efficiency: justice is a scarce resource, and econom-

ic rules favour the speedy and effective adjudication of as many simultane-

ous claims as our other priorities (chiefly concerns over individual victim

compensation and litigative autonomy4) will permit. There is also a strong

argument that statistical evidence of almost every factor in a single, aggre-

gate trial will be more accurate “proof” than any number of individual

claims, pursued on a balance of probabilities, and in fact the awards even

in individual trials could similarly benefit.5Indeed, this seems self-evident

on every issue with the exception of the eventual distribution of the award,

which is of necessity an individualistic exercise.6

In Canadian discrimination law, as in the United States,7statistical evi-

dence is increasingly employed both with respect to probabilistic character-

istics of employees for the purposes of designing reasonable standards8

2 See generally L. Walker & J. Monahan, “Sampling Damages” (1998) 83 Iowa L. Rev.

545 at 546: “A complete solution of the number problem in mass torts can only be

achieved by … randomly sampling damages without apology.”

3 It has been held, for instance, that statistical proof of damages is particularly appropri-

ate “where the conduct of wrongdoers has rendered it difficult to ascertain the dam-

ages suffered with the precision otherwise possible.”: Blue Cross and Blue Shield of

New Jersey, Inc. v. Phillip Morris, Inc., 133 F. Supp. 2d 162 (E.D.N.Y. 2001) at 169 [here-

inafter Blue Cross], citing New York Pattern Jury Instructions, 2:277 Damages: General

— Commentary (3d 2000).

4 Though there is reason to believe that the individual interest may actually be

enhanced in aggregate claims; see, for instance, D. Hensler, “Resolving Mass Toxic

Torts: Myths and Realities” (1989) U. Ill. L. Rev. 89 at 104: “The use of formal

aggregative procedures may provide more litigant control over the litigation process,

more opportunity for litigant participation in the process, and a better match between

victims” losses and compensation for those losses”; to similar effect see Kaplow &

Shavell, “Fairness vs. Welfare” (2001) 114 Harv. L. Rev. 961 at 1207.

5 See, for instance, J. Koeler & D. Shaviro, “Veridical Verdicts: Increasing Verdict Accu-

racy Through the Use of Overtly Probabilistic Evidence and Methods” (1990) 75 Cor-

nell L. Rev. 247.

6 See the discussion below at notes 21 through 26 and accompanying text.

7 See International Brotherhood of Teamsters v. United States, 431 U.S. 324 (1977) at 329:

“[O]ur cases make it unmistakably clear that ‘statistical analyses have served and will

continue to serve an important role in cases in which the existence of discrimination

is a disputed issue.”

8Ontario Human Rights Commission v. Etobicoke,[1982] 1 S.C.R. 202 at 212: “It seems to

me, however, that in cases such as this, statistical and medical evidence based upon

observation and research on the question of aging, if not in all cases absolutely neces-

sary, will certainly be more persuasive than the testimony of persons, albeit with great

experience in fire-fighting, to the effect that fire-fighting is “a young man’s game.”